Comparative Analysis To Determine Predictive Model Accuracy : A dynamic currency exchange rate predictive model development using SAP HANA Predictive Analytic Library (PAL) algorithm
Oke, Mudiaga (2014)
Oke, Mudiaga
Metropolia Ammattikorkeakoulu
2014
All rights reserved
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-201404073938
https://urn.fi/URN:NBN:fi:amk-201404073938
Tiivistelmä
The present thesis describes the development and implementation of a dynamic currency exchange rate predictive model. The aim of the thesis was to measure and determine the accuracy of a dynamic currency exchange rate predictive model by analysing different historical data samples.
The theoretical framework of the thesis focused on research into different disciplines related to predicted analytics and the different data mining algorithms. The study was carried out using quantitative data samples and SAP high performance analytic appliance predictive analysis library (PAL) Time series double exponential algorithm. The measurement was done by comparing the predicted or forecasted exchange rates against the actual exchange rates. Standard statistical methods were used to determine the accuracy of the predictive model.
The results of the study showed that last three months data sample or most recent data gives better predictive results for short term forecasting while the full data sample or entire data set gives better result for longer term forecasting.
Based on the study, it is recommended that fundamental analysis of currency exchange method which takes account of the driving forces behind currency exchange rates such as political and economic situation, the rise and fall of interest rates and other economic indicators should be incorporated along technical analysis which involves the use of historical data to get give better accuracy.
The theoretical framework of the thesis focused on research into different disciplines related to predicted analytics and the different data mining algorithms. The study was carried out using quantitative data samples and SAP high performance analytic appliance predictive analysis library (PAL) Time series double exponential algorithm. The measurement was done by comparing the predicted or forecasted exchange rates against the actual exchange rates. Standard statistical methods were used to determine the accuracy of the predictive model.
The results of the study showed that last three months data sample or most recent data gives better predictive results for short term forecasting while the full data sample or entire data set gives better result for longer term forecasting.
Based on the study, it is recommended that fundamental analysis of currency exchange method which takes account of the driving forces behind currency exchange rates such as political and economic situation, the rise and fall of interest rates and other economic indicators should be incorporated along technical analysis which involves the use of historical data to get give better accuracy.